Metabolomics analysis of differential chemical constituents and α-glucosidase inhibiting activity of Phyllanthus urinaria L. root, stem, leaf and fruit

Abstract The differential chemical constituents of the different Phyllanthus urinaria L. (PUL) parts were investigated by UPLC/Q-TOF MS-based metabolomics. A total of 69 compounds were tentatively identified in the whole plant extract and 35 of them were common to root, stem, leaf and fruit parts. And four compounds were selected as biomarkers for leaves, fruits, stems and roots, respectively. The four PUL parts all had good α-glucosidase inhibitory activities and the activities of fruit, root and stem extracts were about fivefold higher than the leaf part. The hierarchical cluster analysis and heat map were used to explore the relationship between the α-glucosidase inhibitory activities and chemical constituent differences of four PUL parts. Graphical Abstract


Introduction
Phyllanthus urinaria L. (PUL) is belonged to the Euphorbiaceae family and is widely grown in tropical and subtropical regions of Asian countries (Mao et al. 2016). Traditionally, the PUL is used for the treatment of various complications in different regions of the world (Geethangili and Ding 2018). The commercial PUL capsules and tablets have been successfully applied to the clinical treatment of chronic hepatitis B. Many constituents are reported in PUL (Geethangili and Ding 2018). These constituents distribute in different plant parts which hence result in bioactivity differences (Ueda et al. 2000). However, the detailed investigation around the different PUL parts is still not clear.
In this study, the chemical constituents of the roots, stems, leaves and fruits of PUL were carried out by UPLC/Q-TOF MS-based metabolomics. The principal component analysis (PCA) and orthogonal partial least squares discriminant analysis (OPLS-DA) were used to reduce dimensionality and determine biomarkers of differences in each plant part. The a-glucosidase inhibitory activities of different PUL parts were evaluated. The relationship between the a-glucosidase inhibitory activities and chemical constituent differences of four PUL parts was explained. Till now, the comprehensive profile for the compositional differences and a-glucosidase inhibitory activity differences of the roots, stems, leaves and fruits of PUL is not reported in the literature.

Results and discussion
2.1. Qualitative analysis of chemical constituents of four PUL parts PUL was collected in Hangzhou city, Zhejiang Province, China in October 2020. These plants were authenticated by Dr. Chu Chu (Zhejiang University of Technology, Hangzhou, China). The UPLC/Q-TOF MS approach with negative ion mode was utilised to acquire comprehensive chemical constituent information. The qualitative analysis results of the plant extracts were summarised in Supplementary Material Table S1. According to the obtained high resolution MS data and reference substances, 69 compounds including 28 tannins were tentatively identified (Chen et al. 1999). And 35 of them were observed to be all present in the four plant parts. The representative base peak chromatogram (BPC) for PUL whole plant extract was shown in Supplementary Material Figure S1, the BPC for the roots, stems, leaves and fruits of PUL were shown in Supplementary Material Figure S2.

The differential chemical constituents analysis of four PUL parts
The QC samples clustered into one small set near the coordinate origin of the PCA score plot (Supplementary Material Figure S3a), demonstrating that the UPLC/Q-TOF MS method was stable and reproducible. The PCA score plot demonstrated that the metabolomics data of the roots, stems, leaves and fruits of PUL was well clustered within the group. The PCA principal component number was set to 4, which had good reproducibility (R 2 X ¼ 0.792) and predictability (Q 2 ¼0.693). Supplementary Material Figure S3b showed the hierarchical cluster analysis (HCA) of the metabolomics data of the roots, stems, leaves and fruits of PUL. The results showed that the metabolites of the roots, stems, leaves and fruits of PUL were significantly different.
As shown in Supplementary Material Figure S4, the OPLS-DA score plots were on the left, and the permutation test results of the OPLS-DA models were on the right. All observed values were within the Hotelling's T2 (95%) ellipse, indicating that there were no obvious outliers. As displayed in Supplementary Material Table S2, the parameter R 2 Y represented the level of agreement between the sample and the model, and the Q 2 of the score plot indicated the prediction accuracy of the model. 200 permutation tests were performed on the overfitting of the OPLS-DA model. R2 and Q2 represented the vertical intercept of the regression lines, and the data points were lower than the original point on the right, indicating that the model was valid. The results showed that the OPLS-DA models had good predictive ability and could explain the differences between samples. In the S-plot of OPLS-DA (Supplementary Material Figure S5), the farther the data points were from the center, the more they contributed to the difference between groups, and the greater the possibility of becoming differential metabolites. According to the ranking of VIP values in the six groups of OPLS-DA models, metabolites with VIP values greater than 5 were selected as differential chemical constituents, and a total of 29 differential metabolites were identified.
The heat map (Supplementary Material Figure S6) was used to indicate the distribution of these differential metabolites among the roots, stems, leaves and fruits of PUL. The value of the mass spectrum signal intensity (peak height) was taken logarithmically and used as the value of the heat map. The heat map was centrally processed. According to the heat map, the relative amount of valoneic acid dilactone was the largest in the leaf, so valoneic acid dilactone was considered as a biomarker of the leaf part of PUL. Similarly, epigallocatechin-(4beta->8)-catechin, quercetin 3-(4 0 '-acetylrhamnoside) 7-rhamnoside and Boehmenan were assigned as the biomarkers of fruit, stem and root, respectively.

In vitro a-glucosidase inhibitory assay of PUL extracts
The extracts from roots, stems, leaves, fruits and the whole plant of PUL were assessed for their inhibitory activities against a-glucosidase. The IC 50 values of fruit, root, stem, leaf parts and whole plant extract were 0.47 ± 0.01, 0.67 ± 0.05, 1.03 ± 0.05, 5.62 ± 0.68 and 0.72 ± 0.08 lg/mL, respectively. The IC 50 value of acarbose was determined to be 857.4 ± 61.1 lg/mL under the same conditions (Lopez-Angulo et al. 2022). As shown by the dose-response curves (Supplementary Material Figure S7), all plant extracts showed better inhibitory activities than the positive control acarbose. The a-glucosidase inhibitory activities of fruit, root and stem parts were higher than leaf part, which was consistent with the previously obtained HCA result. From the HCA result of Supplementary Material Figure S3b, the fruit, root and stem were divided into a large group (lower distance), while the leaf part was individually grouped into one group (larger distance). According to the heat map, the relative contents of epigallocatechin-(4beta->8)-catechin, epigallocatechin, isostrictinin, corilagin, repandusinic acid A and phyllanthusiin A in fruit, root and stem were significantly higher than those in leaf part. These six compounds were identified as the potential active compounds most closely related to the a-glucosidase inhibitory activity of PUL. Among them, corilagin and repandusinic acid A have been reported their a-glucosidase inhibitory activities in PUL with IC 50 1.70 ± 0.03 and 6.10 ± 0.10 lM, respectively (Trinh et al. 2016). Isostrictinin was one of the effective constituents of Psidium guajava for treating diabetes by improving insulin sensitivity (Patel et al. 2012).

Conclusions
The phytochemical profiles and a-glucosidase inhibitor activities of the roots, stems, leaves and fruits of PUL were investigated. This study demonstrated that the roots, stems, leaves and fruits of PUL all have good a-glucosidase inhibitory activities, and the inhibitory activity differences are related to the chemical constituent differences of different plant parts.

Disclosure statement
No potential conflict of interest was reported by the authors.

Funding
This work was supported by the cooperative project with Zhejiang Hisoar Pharmaceutical Co., Ltd. (KYY-HX-20180525).